PLoS ONE (Jan 2013)

Association between reduction of plasma adiponectin levels and risk of bacterial infection after gastric cancer surgery.

  • Hiroshi Yamamoto,
  • Kazuhisa Maeda,
  • Yoshitaka Uji,
  • Hiroshi Tsuchihashi,
  • Tsuyoshi Mori,
  • Tomoharu Shimizu,
  • Yoshihiro Endo,
  • Aya Kadota,
  • Katsuyuki Miura,
  • Yusuke Koga,
  • Toshinori Ito,
  • Tohru Tani

DOI
https://doi.org/10.1371/journal.pone.0056129
Journal volume & issue
Vol. 8, no. 3
p. e56129

Abstract

Read online

BACKGROUND AND PURPOSE: Infections are important causes of postoperative morbidity after gastric surgery; currently, no factors have been identified that can predict postoperative infection. Adiponectin (ADN) mediates energy metabolism and functions as an immunomodulator. Perioperative ADN levels and perioperative immune functioning could be mutually related. Here we evaluated a potential biological marker to reliably predict the incidence of postoperative infections to prevent such comorbidities. METHODS: We analyzed 150 consecutive patients who underwent elective gastric cancer surgery at the Shiga University of Medical Science Hospital (Shiga, Japan) from 1997 to 2009; of these, most surgeries (n = 100) were performed 2008 onwards. The patient characteristics and surgery-related factors between two groups (with and without infection) were compared by the paired t-test and χ(2) test, including preoperative ADN levels, postoperative day 1 ADN levels, and ADN ratio (postoperative ADN levels/preoperative ADN levels) as baseline factors. Logistic regression analysis was performed to access the independent association between ADN ratio and postoperative infection. Finally, receiver operating curves (ROCs) were constructed to examine its clinical utility. RESULTS: Sixty patients (40%) experienced postoperative infections. The baseline values of age, American Society of Anesthesiologists physical status, total operating time, blood loss, surgical procedure, C-reactive protein (CRP) levels, preoperative ADN levels, and ADN ratio were significantly different between groups. Logistic regression analysis using these factors indicated that type 2 diabetes mellitus (T2DM) and ADN ratio were significantly independent variables (*p<0.05, ** p<0.01, respectively). ROC analysis revealed that the useful cutoff values (sensitivity/specificity) for preoperative ADN levels, ADN ratio, blood loss, operating time, and CRP levels were 8.81(0.567/0.568), 0.76 (0.767/0.761), 405 g (0.717/0.693), 342 min (0.617/0.614), and 8.94 mg/dl (0.583/0.591), respectively. CONCLUSION: T2DM and ADN ratio were independent predictors of postoperative infection and ADN ratio was the most useful predictor for postoperative infection.